You come off as assuming that the people in this thread are not aware of the personal identity debate. That doesn’t really strike me as productive.
I know the debate exists, I just think the wrong side is winning (in this little corner of the Internet).
These discussions usually occur in an atmosphere where there is far more presumption than there is knowledge, regarding the relationship between the physical brain and elements of personhood like mind, consciousness, or identity.
The default attitude is that a neuron is just a sort of self-healing transistor, that the physical-computational reality in the brain that is relevant for the existence of a person is a set of trillions of physically distinct but causally connected elementary acts of information processing, that the person exists in or alongside these events in some vague way that is not completely specified or understood, and that so long as the cloud of information processing continues in a vaguely similar way, or so long as it is instantiated in a way with vaguely similar causality, the person will continue to exist or will exist again, thanks to the vague and not-understood principle of association that links physical computation and self.
The view of the self which naturally arises subjectively is that it is real and that it persists in time. But because of the computational atomism present in the default attitude (described in the previous paragraph), and also because of MWI dogma and various thought-experiments involving computer programs, the dominant tendency is to say, the natural view is naive and wrong, there’s no continuity, you are as much your copies and your duplicates elsewhere in the multiverse as you are this-you, and so on ad infinitum, in a large number of permutations on the theme that you aren’t what you think you are or what you appear to be.
Another reason that people are willing to sacrifice the natural view wholesale has to do with attachment to the clarity that comes from mathematical, physical, and computational concepts. These concepts also have an empirical side to their attraction, in that science has validated their use to a certain degree. But I think there is also a love of the types of objective thought that we already know how to perform, a love of them for their own sake, even when they do not easily mesh with something we would otherwise consider a reality. An example would be the flow of time. If there were no mathematical physics which treats time as just another sort of space, we would say, of course time passes; it may be a little mysterious, a little hard to describe (see St Augustine), but of course it’s real. But since we have geometric models of space-time which are most easily apprehended in a timeless or static fashion (as a single self-existent geometric object), the flow of time is relegated to consciousness, subjectivity, illusion, even denied outright.
The matter-of-fact belief in uploading (in the sense that a digital emulation of your brain will be conscious and will be you), in cryonics, and a few of the other less baroque permutations of identity, doesn’t necessarily derive from this quasi-platonic absorption in abstract objectivity. It can also come from “thought-experiment common-sense”, as in your scenario of a world where cryonics or uploading is already common. But the abstract attachment to certain modes of thought is a significant factor in this particular intellectual environment, and in discussions where there is an attempt to think about these matters from first principles. And apparently it needs to be pointed out that everything which is “subjective”, and which presents a problem for the standard picture (by which I mean, the combination of computational atomism and physical ontology), is quickly discarded as unreal or is treated as a problem that will take care of itself somehow, even though the standard picture is incredibly vague regarding how identity, mind, or consciousness relates to the physical and biological description.
I lean in the other direction, because the standard picture is still very vague, and also because subjective appearances have evidential value. We have every reason to look for models of reality in which time is real, the self is real, and so on, rather than bravely biting the naturalistic bullet and asserting that all that stuff is somehow unreal or otherwise ontologically secondary. Last year I suggested here that the exact physical correlate of the self might be a specific, very large irreducible tensor factor in the quantum state of the brain. That wasn’t a very popular idea. There’s also the idea that the self is a little more mesoscopic than that, and should be identified with a dynamic “functional organization” glued together more by causality than by anything else. I think this idea is a little fuzzy, a little problematic, but it’s still superior to the patternist theories according to which physical continuity in space and time has nothing to do with identity; and perhaps you can see that on this theory also, a destructive upload or a cryonic resurrectee isn’t you—because you were the particular dynamical process which terminated when you died, and the upload and the cryo-copy are distinct processes, with a definite beginning in time which came long after the end of the earlier process on which they were modeled.
You ask how, in a world where cryonic suspension and mind uploading are commonplace, a person could arrive at the intuition that identity does not persist across these transformations. All it would take is the knowledge that a self is an entity of type X, and that in these transformations, one such entity is destroyed and another created. These questions won’t remain unanswered forever, and I see no reason to think that the final answers will be friendly to or consistent with a lax attitude towards personal identity. To be is to be something, and inside appearances already tell us that each of us is one particular thing which persists in time. All that remains is to figure out what that thing is, from the perspective of natural science.
Thank you for the thoughtful reply. I see the problem with overly cavalier attitudes to personal identity as a reproducible pattern, but I’m not quite willing to let the continuous process view out of the hook that easily either.
It seems awfully convenient that your posited process of personal identity survives exactly those events, blinking ones eyes, epileptic fits, sleep, coma, which are not assumed to disrupt personal identity in everyday thought. Unless we have some actual neuroscience to point to, all I know is that I’m conscious right now. I don’t see why I should assume that whatever process creates my conscious feeling of self is tied exactly to the layer of physical metabolism. It could be dissolved and created anew several times a second, or it could vanish whenever I lose consciousness. I (or the new me) would be none the wiser, since the consciousness going away or coming into being doesn’t feel like anything by itself. Assuming that the continuity of a physical system is indeed vital, it could be that it’s tied to cellular metabolism in exactly the convenient way, but I’m not buying an argument that seems to basically come down to an appeal to common sense.
This is also why I’m a bit wary of your answer to the thought experiment. I’m not entirely sure how the process of discovery you describe would happen. Suppose that people today do neuroscience, and identify properties that seem to always be present in the brains of awake, alert and therefore supposedly conscious people. These properties vanish when the people lose consciousness. Most likely scientists would not conclude that the dissolution of this state intrinsically tied to consciousness means that the subject’s personal identity is gone, since common-sense reason assures us that we retain our personal identity through unconsciousness. I don’t see any way of actually knowing this though. Going to sleep and waking up would feel exactly the same for the sleep-goer and up-waker if the unconsciousness caused a destruction and reconstruction of personal identity than if it would not. I assume that people living in a society with ubiquitous revival from zero metabolism would have similar preconceptions about the revival. The situation is of course likely to be different once we have a better understanding of exactly how the brain works, but lacking that understanding, I’m having some trouble envisioning exactly how the destruction of personal identity could be determined to be intractably tied to the observed entity X.
Finally there’s the question of exactly what it means for a physical system to be intrinsically tied to being continuous in space-time. I can’t think of any phenomenon in classical mechanics where I could point to any property of the system that would be disrupted if the system got disassembled and reassembled mid-evolution. There may be something like that in quantum physics though, I don’t have much intuition regarding that.
It seems awfully convenient that your posited process of personal identity survives exactly those events, blinking ones eyes, epileptic fits, sleep, coma, which are not assumed to disrupt personal identity in everyday thought.
The philosophical habit of skeptically deconstructing basic appearances seems to prepare people badly for the task of scientifically understanding consciousness. When considering the relationship between mind and matter, it’s a little peculiar to immediately jump to complicated possibilities (“whatever process creates my conscious feeling of self … could be dissolved and created anew several times a second”) or to the possibility that appearances are radically misleading (consciousness might be constantly “going away or coming into being” without any impact on the apparent continuity of experience or of personal existence). Just because there might be an elephant around the next corner doesn’t mean we should attach much significance to the possibility.
I’m not entirely sure how the process of discovery you describe would happen… The situation is of course likely to be different once we have a better understanding of exactly how the brain works, but lacking that understanding, I’m having some trouble envisioning exactly how the destruction of personal identity could be determined to be intractably tied to the observed entity X.
It is unlikely that society would develop the capacity for mind uploading and cryonic resurrection without also coming to understand, very thoroughly, how the brain works. We may think we can imagine these procedures being performed locally in the brain, with the global result being achieved by brute force, without a systemic understanding. But to upload or reanimate you do have to know how to put the pieces back together, and the ability to perform local reassembly of parts correctly, in a physical or computational sense, also implies some ability to perform local reassembly conceptually.
In fact it would be reasonable to argue that without a systemic understanding, attempts at uploading and cryonic restoration would be a game of trial and error, producing biased copies which deviate from their originals in unpredictable ways. Suppose you use high-resolution fMRI time series to develop state-machine simulations of microscopic volumes in the brain of your subject (each such “voxel” consisting of a few hundred neighboring neurons). You will be developing a causal model of the parts of the subject’s brain by analysing the time series. It’s easy to imagine the analysis assuming that interactions only occur between neighboring voxels, or even next-nearest neighbors, and thereby overlooking long-range interactions due to long axonal fibers. The resulting upload will have lost some of the causal structure of its prototype.
The possibility of elementary errors like this, to say nothing of whatever more subtle mistakes may occur, implies that we can’t really trust procedures like this without simultaneously developing that “better understanding of exactly how the brain works”.
I can’t think of any phenomenon in classical mechanics where I could point to any property of the system that would be disrupted if the system got disassembled and reassembled mid-evolution.
How about the property of being an asymptotically bound system, in the absence of active disassembly by external forces? To me that still seems way too weak to be the ontological basis of physical identity, but that is (more or less) the philosopher Mario Bunge’s definition of systemhood. (Btw, Bunge was a physicist before he was a philosopher.)
The philosophical habit of skeptically deconstructing basic appearances seems to prepare people badly for the task of scientifically understanding consciousness. When considering the relationship between mind and matter, it’s a little peculiar to immediately jump to complicated possibilities
It wasn’t philosophers who came up with general relativity and quantum mechanics when everyday intuition about nature didn’t quite add up in some obscure corner cases. Coming up with a simple model that seems to resolve contradictions even if it doesn’t quite fit everyday intuition seems to be useful in gaining a better understanding of things.
I’m also having genuine difficulties going anywhere past the everyday intuition with the idea of the discontinuity of personal identity separate from the discontinuity of mindstate. The idea of there being only a sequence of conscious moments instead of an intrinsic continuity doesn’t present any immediately obvious contradiction and doesn’t have the confusion of exactly what the mindstate-independent component of continuous personal is really supposed to be.
Of course going with the mindstate history view, now the difference becomes the sliding scale of possible differences from the previous state. It looks like personal continuity would become a matter of degree rather than a binary thing, which pushes things further into the unintuitive.
I can’t think of any phenomenon in classical mechanics where I could point to any property of the system that would be disrupted if the system got disassembled and reassembled mid-evolution.
How about the property of being an asymptotically bound system, in the absence of active disassembly by external forces?
I’m afraid I don’t understand what that means. Can you give more concrete examples of physical things that do or don’t have this property?
The idea of there being only a sequence of conscious moments instead of an intrinsic continuity doesn’t present any immediately obvious contradiction
It contradicts the experience of time passing—the experience of change. The passage of time is an appearance, and an appearance is something stronger than an intuition. An intuition is a sort of guess about the truth, and may or may not be true, but one normally supposes that appearances definitely exist, at least as appearances. The object implied by a hallucination may not exist, but the hallucination itself does exist. It is always a very radical move to assert that an alleged appearance does not exist even on the plane of appearance. When you deny the existence of a subject which persists in time and which experiences time during that persistent existence, you are right on the edge of denying a fundamental appearance, or perhaps over the edge already.
Normally one supposes that there is an elemental experience of time flowing, and that this experience itself exists in time and somehow endures through time. When you disintegrate temporal experience into a set of distinct momentary experiences not actually joined by temporal flow, the most you can do to retain the appearance of flow is to say that each momentary experience has an illusion of flow. Nothing is ever actually happening in consciousness, but it always looks like it is. Consciousness in every moment is a static thing, but it always has an illusion of change embedded in it. (I suppose you could have a wacky theory of dynamic momentary experiences, whereby they’re all still distinct, but they do come into and then go out of existence, and the momentary appearance of flow is somehow derived from this; the illusion would then be the illusion of persistent flow.)
To sum up, it’s hard to have an actual experience of persistent flow without actually persisting. If you deny that, then either the experience of persistence or the experience of flow has to be called an illusion. And if one becomes willing to assert the persistence of the perceiver, the one having the experience, then there’s no particular reason to be minimalist about it—which I think would be the next step up for someone retreating from a position of temporal atomism. “OK, when I’m aware that time is passing, maybe it’s likely that I persistently exist throughout that experience. But what about when I’m just in the moment, and there’s a gap in time before I contrast the present with the past via memory? How do I know that there was continuity?” The simplest interpretation of this is to say that there was continuity, but you weren’t paying attention to it.
How about the property of being an asymptotically bound system, in the absence of active disassembly by external forces?
I’m afraid I don’t understand what that means. Can you give more concrete examples of physical things that do or don’t have this property?
Consider two gravitating objects. If they orbit a common center of gravity forever, we can call that asymptotically bound; if they eventually fly apart and become arbitrarily distant, they are asymptotically free. You could start with a system which, in the absence of perturbing influences, is asymptotically bound; then perturb it until it became asymptotically free, and then perturb it again in order to restore asymptotic boundedness.
I know the debate exists, I just think the wrong side is winning (in this little corner of the Internet).
These discussions usually occur in an atmosphere where there is far more presumption than there is knowledge, regarding the relationship between the physical brain and elements of personhood like mind, consciousness, or identity.
The default attitude is that a neuron is just a sort of self-healing transistor, that the physical-computational reality in the brain that is relevant for the existence of a person is a set of trillions of physically distinct but causally connected elementary acts of information processing, that the person exists in or alongside these events in some vague way that is not completely specified or understood, and that so long as the cloud of information processing continues in a vaguely similar way, or so long as it is instantiated in a way with vaguely similar causality, the person will continue to exist or will exist again, thanks to the vague and not-understood principle of association that links physical computation and self.
The view of the self which naturally arises subjectively is that it is real and that it persists in time. But because of the computational atomism present in the default attitude (described in the previous paragraph), and also because of MWI dogma and various thought-experiments involving computer programs, the dominant tendency is to say, the natural view is naive and wrong, there’s no continuity, you are as much your copies and your duplicates elsewhere in the multiverse as you are this-you, and so on ad infinitum, in a large number of permutations on the theme that you aren’t what you think you are or what you appear to be.
Another reason that people are willing to sacrifice the natural view wholesale has to do with attachment to the clarity that comes from mathematical, physical, and computational concepts. These concepts also have an empirical side to their attraction, in that science has validated their use to a certain degree. But I think there is also a love of the types of objective thought that we already know how to perform, a love of them for their own sake, even when they do not easily mesh with something we would otherwise consider a reality. An example would be the flow of time. If there were no mathematical physics which treats time as just another sort of space, we would say, of course time passes; it may be a little mysterious, a little hard to describe (see St Augustine), but of course it’s real. But since we have geometric models of space-time which are most easily apprehended in a timeless or static fashion (as a single self-existent geometric object), the flow of time is relegated to consciousness, subjectivity, illusion, even denied outright.
The matter-of-fact belief in uploading (in the sense that a digital emulation of your brain will be conscious and will be you), in cryonics, and a few of the other less baroque permutations of identity, doesn’t necessarily derive from this quasi-platonic absorption in abstract objectivity. It can also come from “thought-experiment common-sense”, as in your scenario of a world where cryonics or uploading is already common. But the abstract attachment to certain modes of thought is a significant factor in this particular intellectual environment, and in discussions where there is an attempt to think about these matters from first principles. And apparently it needs to be pointed out that everything which is “subjective”, and which presents a problem for the standard picture (by which I mean, the combination of computational atomism and physical ontology), is quickly discarded as unreal or is treated as a problem that will take care of itself somehow, even though the standard picture is incredibly vague regarding how identity, mind, or consciousness relates to the physical and biological description.
I lean in the other direction, because the standard picture is still very vague, and also because subjective appearances have evidential value. We have every reason to look for models of reality in which time is real, the self is real, and so on, rather than bravely biting the naturalistic bullet and asserting that all that stuff is somehow unreal or otherwise ontologically secondary. Last year I suggested here that the exact physical correlate of the self might be a specific, very large irreducible tensor factor in the quantum state of the brain. That wasn’t a very popular idea. There’s also the idea that the self is a little more mesoscopic than that, and should be identified with a dynamic “functional organization” glued together more by causality than by anything else. I think this idea is a little fuzzy, a little problematic, but it’s still superior to the patternist theories according to which physical continuity in space and time has nothing to do with identity; and perhaps you can see that on this theory also, a destructive upload or a cryonic resurrectee isn’t you—because you were the particular dynamical process which terminated when you died, and the upload and the cryo-copy are distinct processes, with a definite beginning in time which came long after the end of the earlier process on which they were modeled.
You ask how, in a world where cryonic suspension and mind uploading are commonplace, a person could arrive at the intuition that identity does not persist across these transformations. All it would take is the knowledge that a self is an entity of type X, and that in these transformations, one such entity is destroyed and another created. These questions won’t remain unanswered forever, and I see no reason to think that the final answers will be friendly to or consistent with a lax attitude towards personal identity. To be is to be something, and inside appearances already tell us that each of us is one particular thing which persists in time. All that remains is to figure out what that thing is, from the perspective of natural science.
Thank you for the thoughtful reply. I see the problem with overly cavalier attitudes to personal identity as a reproducible pattern, but I’m not quite willing to let the continuous process view out of the hook that easily either.
It seems awfully convenient that your posited process of personal identity survives exactly those events, blinking ones eyes, epileptic fits, sleep, coma, which are not assumed to disrupt personal identity in everyday thought. Unless we have some actual neuroscience to point to, all I know is that I’m conscious right now. I don’t see why I should assume that whatever process creates my conscious feeling of self is tied exactly to the layer of physical metabolism. It could be dissolved and created anew several times a second, or it could vanish whenever I lose consciousness. I (or the new me) would be none the wiser, since the consciousness going away or coming into being doesn’t feel like anything by itself. Assuming that the continuity of a physical system is indeed vital, it could be that it’s tied to cellular metabolism in exactly the convenient way, but I’m not buying an argument that seems to basically come down to an appeal to common sense.
This is also why I’m a bit wary of your answer to the thought experiment. I’m not entirely sure how the process of discovery you describe would happen. Suppose that people today do neuroscience, and identify properties that seem to always be present in the brains of awake, alert and therefore supposedly conscious people. These properties vanish when the people lose consciousness. Most likely scientists would not conclude that the dissolution of this state intrinsically tied to consciousness means that the subject’s personal identity is gone, since common-sense reason assures us that we retain our personal identity through unconsciousness. I don’t see any way of actually knowing this though. Going to sleep and waking up would feel exactly the same for the sleep-goer and up-waker if the unconsciousness caused a destruction and reconstruction of personal identity than if it would not. I assume that people living in a society with ubiquitous revival from zero metabolism would have similar preconceptions about the revival. The situation is of course likely to be different once we have a better understanding of exactly how the brain works, but lacking that understanding, I’m having some trouble envisioning exactly how the destruction of personal identity could be determined to be intractably tied to the observed entity X.
Finally there’s the question of exactly what it means for a physical system to be intrinsically tied to being continuous in space-time. I can’t think of any phenomenon in classical mechanics where I could point to any property of the system that would be disrupted if the system got disassembled and reassembled mid-evolution. There may be something like that in quantum physics though, I don’t have much intuition regarding that.
The philosophical habit of skeptically deconstructing basic appearances seems to prepare people badly for the task of scientifically understanding consciousness. When considering the relationship between mind and matter, it’s a little peculiar to immediately jump to complicated possibilities (“whatever process creates my conscious feeling of self … could be dissolved and created anew several times a second”) or to the possibility that appearances are radically misleading (consciousness might be constantly “going away or coming into being” without any impact on the apparent continuity of experience or of personal existence). Just because there might be an elephant around the next corner doesn’t mean we should attach much significance to the possibility.
It is unlikely that society would develop the capacity for mind uploading and cryonic resurrection without also coming to understand, very thoroughly, how the brain works. We may think we can imagine these procedures being performed locally in the brain, with the global result being achieved by brute force, without a systemic understanding. But to upload or reanimate you do have to know how to put the pieces back together, and the ability to perform local reassembly of parts correctly, in a physical or computational sense, also implies some ability to perform local reassembly conceptually.
In fact it would be reasonable to argue that without a systemic understanding, attempts at uploading and cryonic restoration would be a game of trial and error, producing biased copies which deviate from their originals in unpredictable ways. Suppose you use high-resolution fMRI time series to develop state-machine simulations of microscopic volumes in the brain of your subject (each such “voxel” consisting of a few hundred neighboring neurons). You will be developing a causal model of the parts of the subject’s brain by analysing the time series. It’s easy to imagine the analysis assuming that interactions only occur between neighboring voxels, or even next-nearest neighbors, and thereby overlooking long-range interactions due to long axonal fibers. The resulting upload will have lost some of the causal structure of its prototype.
The possibility of elementary errors like this, to say nothing of whatever more subtle mistakes may occur, implies that we can’t really trust procedures like this without simultaneously developing that “better understanding of exactly how the brain works”.
How about the property of being an asymptotically bound system, in the absence of active disassembly by external forces? To me that still seems way too weak to be the ontological basis of physical identity, but that is (more or less) the philosopher Mario Bunge’s definition of systemhood. (Btw, Bunge was a physicist before he was a philosopher.)
It wasn’t philosophers who came up with general relativity and quantum mechanics when everyday intuition about nature didn’t quite add up in some obscure corner cases. Coming up with a simple model that seems to resolve contradictions even if it doesn’t quite fit everyday intuition seems to be useful in gaining a better understanding of things.
I’m also having genuine difficulties going anywhere past the everyday intuition with the idea of the discontinuity of personal identity separate from the discontinuity of mindstate. The idea of there being only a sequence of conscious moments instead of an intrinsic continuity doesn’t present any immediately obvious contradiction and doesn’t have the confusion of exactly what the mindstate-independent component of continuous personal is really supposed to be.
Of course going with the mindstate history view, now the difference becomes the sliding scale of possible differences from the previous state. It looks like personal continuity would become a matter of degree rather than a binary thing, which pushes things further into the unintuitive.
I’m afraid I don’t understand what that means. Can you give more concrete examples of physical things that do or don’t have this property?
It contradicts the experience of time passing—the experience of change. The passage of time is an appearance, and an appearance is something stronger than an intuition. An intuition is a sort of guess about the truth, and may or may not be true, but one normally supposes that appearances definitely exist, at least as appearances. The object implied by a hallucination may not exist, but the hallucination itself does exist. It is always a very radical move to assert that an alleged appearance does not exist even on the plane of appearance. When you deny the existence of a subject which persists in time and which experiences time during that persistent existence, you are right on the edge of denying a fundamental appearance, or perhaps over the edge already.
Normally one supposes that there is an elemental experience of time flowing, and that this experience itself exists in time and somehow endures through time. When you disintegrate temporal experience into a set of distinct momentary experiences not actually joined by temporal flow, the most you can do to retain the appearance of flow is to say that each momentary experience has an illusion of flow. Nothing is ever actually happening in consciousness, but it always looks like it is. Consciousness in every moment is a static thing, but it always has an illusion of change embedded in it. (I suppose you could have a wacky theory of dynamic momentary experiences, whereby they’re all still distinct, but they do come into and then go out of existence, and the momentary appearance of flow is somehow derived from this; the illusion would then be the illusion of persistent flow.)
To sum up, it’s hard to have an actual experience of persistent flow without actually persisting. If you deny that, then either the experience of persistence or the experience of flow has to be called an illusion. And if one becomes willing to assert the persistence of the perceiver, the one having the experience, then there’s no particular reason to be minimalist about it—which I think would be the next step up for someone retreating from a position of temporal atomism. “OK, when I’m aware that time is passing, maybe it’s likely that I persistently exist throughout that experience. But what about when I’m just in the moment, and there’s a gap in time before I contrast the present with the past via memory? How do I know that there was continuity?” The simplest interpretation of this is to say that there was continuity, but you weren’t paying attention to it.
Consider two gravitating objects. If they orbit a common center of gravity forever, we can call that asymptotically bound; if they eventually fly apart and become arbitrarily distant, they are asymptotically free. You could start with a system which, in the absence of perturbing influences, is asymptotically bound; then perturb it until it became asymptotically free, and then perturb it again in order to restore asymptotic boundedness.
I’m curious: do you consider sleeping, or falling unconscious after hitting your head, to be as deadly as cryonics?
No. See this discussion, including the spinoff discussion with a now-invisible Roko.
Beautifully put. Thank you.